nside out? No, my cap is not inside out, you bastards. This is the official Coffee with Scott Adams hat. There you go. There you go. You see it's me and then that's also me. This I know this is going to be confusing for you. This is me and this is me. Try to keep those straight.

All right. Did you know that caffeine's impact on your brain could reduce your cravings for alcohol? That's right. Is there anything that coffee can't do? Let me check my notes. Anything coffee can't do? It can't help Kamala Harris win the presidency, but besides that there's nothing it can't do. So coffee, according to a post, might make it easier to get off of alcohol if you wanted to.

Did you know that climate models, as incredible and valuable and accurate as they are, all hundred of them that are different, they're going to make them even better? Wow. Talk about the Golden Age, huh? You got your climate models that are totally believable and now they're going to add, wait for it, AI. That's right. That's right. The existing climate models, oh, almost perfect, almost just right. But what they need is one extra layer of total bullshit called AI. Because if you can't trust your AI model for predicting the temperature of the Earth in 80 years, what can you trust?

Well, for those of you who've never been around predictive models, let me tell you something. It's not the data and it's not the model that determines the answer. Do you know what determines the answer? It's not the data and not the model. What is it? It's the assumptions. Because the assumptions are not data. They're assumptions about, well, I think we got all the variables. And then you compare it to all the other models and you go, well, they're making different assumptions about the clouds, but we like our assumptions. It's the assumptions that give you the output. And if they didn't like the output, they wouldn't tell you what AI did to it. It's pretty much just the assumptions.

Meanwhile, scientists have developed a breakthrough nasal spray that could delay Alzheimer's for over a decade according to SciTechDaily. But they have only put it into animals so far. Now I don't know how many animals actually get Alzheimer's or how many animals even live 10 years, so you can check. But the way you do it is you give it to the animal, let's say a cow, and then you wait 10 years and you say to the cow, hey, do you recognize me? And if a cow just stares at you like it doesn't even know you, that's Alzheimer's. But okay, I'm just kidding. They don't give it to cows. Probably might be a rat or something.

All right. But here's my scientific knowledge question for you. I asked a version of this on X. This will sound like I'm leading to something but it's really for your general knowledge. Number one, if something is tested in the lab and it works in the lab and then it also works in animals, what are the odds it will work in a person? Go in the comments, tell me the odds. There's something that works in the lab and it kills things in the test tube and it works in an animal and it's safe enough and effective enough in the animal. What are the odds that that will translate to human usefulness? Your numbers are way too big. 25%? But you're being funny. The real answer is 5 to 10%. 5 to 10%.

Now let's make it more interesting. There are two reasons that a drug would not be approved for human use. One is if it's too dangerous. Two, if it's not, let's say, bioavailable, meaning that if it could get to the right part of your body, you know, cross whatever membranes and get into whatever parts, then it might be effective. But if it doesn't in a human but maybe it does in a rat, then that would be another reason it doesn't work. Now there could be lots of other reasons it doesn't work, but those are two big ones.

So now I'm going to revise my